A Multi-Armed Bandit Model Selection for Cold-Start User Recommendation

@article{Felcio2017AMB,
  title={A Multi-Armed Bandit Model Selection for Cold-Start User Recommendation},
  author={Cr{\'i}cia Z. Fel{\'i}cio and Kl{\'e}risson Vin{\'i}cius Ribeiro Paix{\~a}o and C{\'e}lia A. Zorzo Barcelos and Philippe Preux},
  journal={Proceedings of the 25th Conference on User Modeling, Adaptation and Personalization},
  year={2017}
}
How can we effectively recommend items to a user about whom we have no information? This is the problem we focus on in this paper, known as the cold-start problem. In most existing works, the cold-start problem is handled through the use of many kinds of information available about the user. However, what happens if we do not have any information? Recommender systems usually keep a substantial amount of prediction models that are available for analysis. Moreover, recommendations to new users… Expand
A contextual approach to improve the user's experience in interactive recommendation systems
TLDR
A new contextual approach to initialize the bandit models is proposed to address both exploration and exploitation goals since the first recommendations by mixing entropic and popular items, which maximizes the user's satisfaction in the long-term run. Expand
Assessing and improving recommender systems to deal with user cold-start problem
Recommender systems are in our everyday life. The recommendation methods have as main purpose to predict preferences for new items based on user’s past preferences. The research related to this topicExpand
A Bandit-Based Ensemble Framework for Exploration/Exploitation of Diverse Recommendation Components
This work presents an extension of Thompson Sampling bandit policy for orchestrating the collection of base recommendation algorithms for e-commerce. We focus on the problem of item-to-itemExpand
Cluster-Based Bandits: Fast Cold-Start for Recommender System New Users
TLDR
This paper introduces a new type of online learning algorithm, cluster-based bandits, that exploits the fact that users can often be grouped into clusters based on the similarity of their preferences, and this allows accelerated learning of new user preferences. Expand
Efficient Context-Aware Sequential Recommender System
TLDR
This work proposes to store the covariance matrix of the previously seen contexts in a compressed space, without losing too much in terms of recommendation quality in the era of big-data content features. Expand
The impact of first recommendations based on exploration or exploitation approaches in recommender systems' learning
TLDR
Experimental results show that the combination of exp-exp at first user's interactions improves the system's effectiveness without adding more effort to users. Expand
Proposed Recommender System For Solving Cold Start Issue Using k-means Clustering and Reinforcement Learning Agent
The cold start problem for new users is making a real challenge in the recommender system's operation to provide suggestions for a new user. This paper suggests a reinforcement learning recommenderExpand
Burst-induced Multi-Armed Bandit for Learning Recommendation
TLDR
A non-stationary and context-free Multi-Armed Bandit problem and a novel algorithm (which is referred to as BMAB) to solve it, which model the reward distribution as a consequence of variations in the intensity of the activity and assist the exploration/exploitation dilemma by exploring the temporal dynamics of the audience. Expand
A keyword-aware recommender system using implicit feedback on Hadoop
TLDR
Experimental results obtained from the analysis of keyword similarity and rating accuracy represent that the proposed architecture of keyword-aware recommenders is effective and promising, as well offer profitable recommendation services on improving the recommendation accuracy. Expand
A Semi-Personalized System for User Cold Start Recommendation on Music Streaming Apps
TLDR
This paper presents a semi-personalized recommendation strategy, based on a deep neural network architecture and on a clustering of users from heterogeneous sources of information, recently deployed on the music streaming service Deezer to address the user cold start problem. Expand
...
1
2
...

References

SHOWING 1-10 OF 61 REFERENCES
Mixing bandits: a recipe for improved cold-start recommendations in a social network
TLDR
This work proposes two novel strategies leveraging neighborhood estimates to improve the learning rate of bandits for cold-start users, extending the well-known UCB1 algorithm and inheriting its asymptotically optimal guarantees. Expand
Preference-Like Score to Cope with Cold-Start User in Recommender Systems
TLDR
This work contributes to this line of work comparing several matrix factorization based SRS under full cold-start user scenario, and proposing a general model selection approach, called ToSocialRec, that leverages existing recommendation models to offer items for new users. Expand
Bandits and Recommender Systems
TLDR
The combination of matrix factorization and bandit algorithms to solve the on-line recommendation problem is driven by considering the recommendation problem as a feedback controlled loop and leads to interactions between the representation learning, and the recommendation policy. Expand
Simultaneous co-clustering and learning to address the cold start problem in recommender systems
TLDR
A hybrid approach is presented that combines collaborative filtering recommendations with demographic information and provides a hybrid recommendation approach that can address the (pure) cold start problem, where no collaborative information is available for new users. Expand
Pairwise preference regression for cold-start recommendation
TLDR
P predictive feature-based regression models are proposed that leverage all available information of users and items, such as user demographic information and item content features, to tackle cold-start problems and scale efficiently as a linear function of the number of observations. Expand
Cold-Start Item and User Recommendation with Decoupled Completion and Transduction
TLDR
This paper proposes a novel algorithmic framework based on matrix factorization that simultaneously exploits the similarity information among users and items to alleviate the cold-start problem and is believed to be the first algorithm that addresses theColdstart problem with provable guarantees. Expand
Multi-level preference regression for cold-start recommendations
TLDR
This work designs a new convex framework for cold-start recommendations, multi-level preference regression (MPR), directly to predict the ratings rather than latent profiles, and demonstrates that the proposed model can achieve the state of the art in the usercold-start scenario and the best performance in other scenarios. Expand
Ensemble contextual bandits for personalized recommendation
TLDR
A meta-bandit paradigm is employed that places a hyper bandit over the base bandits, to explicitly explore/exploit the relative importance of base bandits based on user feedbacks to obtain robust predicted click-through rate (CTR) of web objects. Expand
Cold-Start Recommendation with Provable Guarantees: A Decoupled Approach
TLDR
This paper proposes a novel and general algorithmic framework based on matrix completion that simultaneously exploits the similarity information among users and items to alleviate the cold-start problem and is believed to be the first algorithm that addresses theColdstart problem with provable guarantees on performance. Expand
Switching hybrid for cold-starting context-aware recommender systems
TLDR
A switching hybrid solution that exploits a custom selection of two CARS algorithms, each suited for a particular cold-start situation, and switches between these algorithms depending on the detected recommendation situation (new user, new item or new context). Expand
...
1
2
3
4
5
...